- Email: [email protected]
Documentation
Chapter 18
Documentation This chapter does not include commercial documentation, because every company has its own particular system. The technical documentation should include the items shown in Appendix 5 as the minimum requirement. The end user and vendor should agree on the number of copies for each document. The buyer also has to set up a system to handle vendor documentation. At least the plant engineer, project engineer, and plant foreman should review the documents. A common practice is to make a rubber stamp with a place for each individual to initial the documents. The stamp shall also have the following words: • Approved • Approved as noted • Not approved, resubmit The vendor drawings should have the following stamps on them when received by the buyer: • For approval (when first sent to the client) • Approved for construction • Approved and certified (final accepted documents) The buyer may consider some of the following points when preparing his or her vendor data instruction sheet: • Curves for small, inexpensive pumps or pump packages need only be standard curves, with the curve for the proper impeller trim plotted on the standard curve sheet. Dimensional drawings can also be standard drawings with a certified stamp added. • The review of performance curves, dimensional drawings, and data sheets is the buyer's last chance to correct any mistakes in his or her specifications or in the vendor's final design. After the buyer approves the drawing, the vendor starts fabricating the package. The next step is inspection of the finished product. The inspector will have difficulty rejecting a pump package based on a design error not caught during the drawing approval cycle.
140
Documentation
141
The performance curves should be checked for the following: • That the rated point follows specifications • That the end of the curve performance follows specifications. Usually, the end of the curve should be at least 120 percent of the flow rate at the best efficiency curve. • That the NPSHR is acceptable not only at rated flow, but at maximum and minimum flow rates also • That the shutoff pressure for pumps designed to run in parallel follows specifications. The data sheets should be checked for: • The correct metallurgy. Often, pump vendors substitute specified material for similar material that they regularly stock. Do not, after checking the compatibility of the vendor's standard material, insist on the original material. The vendor will refuse or charge an extraordinary price for pouring a special heat for a few casings and impellers. • The correct mechanical seals or packing. If the pump vendor recommends different seal facing than specified, consult with seal vendors. • The correct couplings. If, for instance, the buyer's specification calls for a Metastream M Series coupling or its equal, and the vendor quotes another brand of coupling, check with the coupling manufacturers for compatibility. • The right size driver. Recalculate the horsepower requirement for the pump's entire flow range and check that the quoted horsepower will meet the requirement, including the specified safety margin. A saving in the safety margin may lead to a smaller frame-size motor. The resulting cost difference may make this vendor the low bidder. Engineering often prepares plant layout and pump foundations based on preliminary dimensional drawings. With ANSI pumps, this is not a problem because these pumps are standardized. However, with larger packaged pumps, revised dimensions may become a problem. Too large a pump may need special highway permits, or the package may not fit into the space allocated to it in the buyer's plant layout. Changed anchor bolt holes may also cause problems. Usually, this only involves the buyer changing his or her foundation drawings, but the foundation may already have been poured and the anchor bolts grouted in place. Chipping out the anchor bolts and regrouting them is not serious, but it is vexing. Try to catch dimensional errors on the vendor drawings. The buyer should also take care regarding paint. Pump manufacturers and packagers like to quote their own manufacturers' standard paint and painting procedures. These may be totally inadequate for the intended package location. For instance, packages placed outdoors on offshore production platforms require three-coat marine paint. Standard manufacturer's paint may peel off even before pump start-up. Insist on the correct paint and painting procedures. It will cost more but will avoid future embarrassment.
•C h a p t e r
19
inspection and Testing General Inspection Inspection procedures may range from a full-blown witnessed performance and N P S H string test of an 8,000 hp boiler feed pump at the pump manufacturer's testing facilities to a cursory visual inspection at the buyer's shop of a 1 hp water pump. Table 19.1 shows a typical inspection checklist that any major company could issue.
Table 19.1 Inspection Requirements for Centrifugal Pumps 1
2
3
b
b
a
a
b
b
C
C
b b
a a
C a
C a
c
a
c
c
Radiograph inspection of impeller Liquid penetrant inspection of impeller Liquid penetrant inspection of casing repairs Impact test Hydrostatic test Performance test Certified performance test and NPSH curve Mechanical run test with vibration measurements Manufacturer's standard mechanical run test Noise level test Dismantling in case of failed mechanical run test Balancing of rotating assembly Hardness test for impeller and casing welds*
Key: I = Centrifugal utility pumps < 50 hp or < 250 psig discharge pressure 2 = All centrifugal pumps > 50 hp or > 250 psig discharge pressure 3 = Process pumps < 50 hp or < 250 psig discharge pressure a = Visual inspection by company's inspector b = Certified records examined by company representative c = Certificates and data supplied by vendor *If HzS is present in fluid
142
Inspection and Testing
143
The buyer must determine the inspection procedures he or she wants. If a witnessed test is necessary, the inspector should be instructed to perform the following tasks: • Visually inspect castings • Check welding procedures • Visually inspect all parts before they are assembled. The castings shall not show signs of repair by peening, plugging, or brazing. • Check running clearances on vendor' s inspection records • Check impeller and rotor balance as recorded on vendor' s inspection records • Check pump package dimensions • Witness tests
Hydrostatic Test Hydrotest the pump casing at 1.5 times its design pressure. The gaskets used for this test should perfectly match the gaskets used in the pump. Leaks past the gaskets are not acceptable. Discard the test gaskets after the test. If any leakage or pressure drop occurs during the test, repeat the test after making necessary repairs. Test pumps whose medium has a specific gravity of 0.7 or less with kerosene. Use dye penetrant solution for the hydrotest if the casing pattern is new or changed, and/or if the pump media will consistently remain above 500*F. Any dyepenetration test shall adhere to ASTM E165. If the buyer does not require a witnessed hydrotest, the vendor shall submit its test data for client review.
Performance Test The buyer may insist on a performance string test, which shall include the actual driver, instrumentation, and eventual gear (see Figure 19.1 for setup). Because the vendor's lack of the electric power required may render this test inconvenient or impossible, most companies allow the vendor to use a shop driver. The shop driver often consists of a hydraulic turbine instead of an electric motor. Preferably, the test shall run at rated speed, but the vendor may run the test at other speeds. In this case, the vendor should submit the conversion-to-rated-speed calculations to the buyer for approval before starting the test. The actual string test will then occur at start-up. The manufacturer usually will use water during the performance test. If the real pump medium is viscous, the manufacturer shall make the necessary correction, according to the latest issue of the Hydraulic Institute Complete Pump Standards. The vendor often submits a list of the test equipment it will use. Upon request, the vendor shall submit calibration data for this equipment. Typical instruments include: • An orifice with a differential pressure gauge to measure the flow rate • Digital tachometer to measure pump rpm
144 •
• • • •
Practical Introduction to Pumping Technology
Double-element watt meter to measure power Pressure gauges to measure pump head Vibration monitor Temperature transmitter and RTD probe Sound-level meter
The performance test shall be combined with the mechanical run test, which shall last a minimum of four hours. The inspector shall check the following data, which should match the actual performance curve: • Head against flow • Power requirements at different flow rates • Efficiency at different flow rates The buyer shall select at least five flow rates for inspection. These should include the following points: • • • • • •
Shut-off Minimum flow as specified by vendor A point between rated and minimum flow Flow at best efficiency point Rated flow 110 percent of rated flow At each point, the inspector shall check the following:
• • • • • • •
Pump shaft vibration Shaft axial displacement Suction pressure Discharge pressure Flow rate Driver input power Noise level
If there is a gear in the pump package, the inspector shall check the gear contact pattern after the conclusion of the performance test. Gear contact shall be consistent. A contact area of less than 75 percent will affect gear performance. After the performance test, it is customary to open multistage and barrel pumps to inspect the rotor assembly. After draining the test liquid, the test crew shall dry all the pump components.
Inspection and Testing
145
ORIFICE
f
ISOL~TION
FLOW
VALVE
VAL
COl~Prl~O
r-
A
TURBINE
CLUTCH
MOTOR
GEAR
PUMP
Figure 19.1 Typical Performance Test Setup
NPSH Test It is a good idea to run a NPSH test on pumps 100 hp and larger where the NPSHA is close to the pump's required NPSH. The suppression test may determine the NPSH at 3 percent differential head loss or at 0 percent loss if the pump is in critical service with variable flow. The NPSH shall be plotted at the following flow rates, as a minimum: • • • •
Minimum continuous flow Midway between minimum flow and rated flow Rated flow 110 percent of rated flow
Documentation This chapter does not include commercial documentation, because every company has its own particular system. The technical documentation should include the items shown in Appendix 5 as the minimum requirement. The end user and vendor should agree on the number of copies for each document. The buyer also has to set up a system to handle vendor documentation. At least the plant engineer, project engineer, and plant foreman should review the documents. A common practice is to make a rubber stamp with a place for each individual to initial the documents. The stamp shall also have the following words: • Approved • Approved as noted • Not approved, resubmit The vendor drawings should have the following stamps on them when received by the buyer: • For approval (when first sent to the client) • Approved for construction • Approved and certified (final accepted documents) The buyer may consider some of the following points when preparing his or her vendor data instruction sheet: • Curves for small, inexpensive pumps or pump packages need only be standard curves, with the curve for the proper impeller trim plotted on the standard curve sheet. Dimensional drawings can also be standard drawings with a certified stamp added. • The review of performance curves, dimensional drawings, and data sheets is the buyer's last chance to correct any mistakes in his or her specifications or in the vendor's final design. After the buyer approves the drawing, the vendor starts fabricating the package. The next step is inspection of the finished product. The inspector will have difficulty rejecting a pump package based on a design error not caught during the drawing approval cycle.
140
Documentation
141
The performance curves should be checked for the following: • That the rated point follows specifications • That the end of the curve performance follows specifications. Usually, the end of the curve should be at least 120 percent of the flow rate at the best efficiency curve. • That the NPSHR is acceptable not only at rated flow, but at maximum and minimum flow rates also • That the shutoff pressure for pumps designed to run in parallel follows specifications. The data sheets should be checked for: • The correct metallurgy. Often, pump vendors substitute specified material for similar material that they regularly stock. Do not, after checking the compatibility of the vendor's standard material, insist on the original material. The vendor will refuse or charge an extraordinary price for pouring a special heat for a few casings and impellers. • The correct mechanical seals or packing. If the pump vendor recommends different seal facing than specified, consult with seal vendors. • The correct couplings. If, for instance, the buyer's specification calls for a Metastream M Series coupling or its equal, and the vendor quotes another brand of coupling, check with the coupling manufacturers for compatibility. • The right size driver. Recalculate the horsepower requirement for the pump's entire flow range and check that the quoted horsepower will meet the requirement, including the specified safety margin. A saving in the safety margin may lead to a smaller frame-size motor. The resulting cost difference may make this vendor the low bidder. Engineering often prepares plant layout and pump foundations based on preliminary dimensional drawings. With ANSI pumps, this is not a problem because these pumps are standardized. However, with larger packaged pumps, revised dimensions may become a problem. Too large a pump may need special highway permits, or the package may not fit into the space allocated to it in the buyer's plant layout. Changed anchor bolt holes may also cause problems. Usually, this only involves the buyer changing his or her foundation drawings, but the foundation may already have been poured and the anchor bolts grouted in place. Chipping out the anchor bolts and regrouting them is not serious, but it is vexing. Try to catch dimensional errors on the vendor drawings. The buyer should also take care regarding paint. Pump manufacturers and packagers like to quote their own manufacturers' standard paint and painting procedures. These may be totally inadequate for the intended package location. For instance, packages placed outdoors on offshore production platforms require three-coat marine paint. Standard manufacturer's paint may peel off even before pump start-up. Insist on the correct paint and painting procedures. It will cost more but will avoid future embarrassment.
•C h a p t e r
19
inspection and Testing General Inspection Inspection procedures may range from a full-blown witnessed performance and N P S H string test of an 8,000 hp boiler feed pump at the pump manufacturer's testing facilities to a cursory visual inspection at the buyer's shop of a 1 hp water pump. Table 19.1 shows a typical inspection checklist that any major company could issue.
Table 19.1 Inspection Requirements for Centrifugal Pumps 1
2
3
b
b
a
a
b
b
C
C
b b
a a
C a
C a
c
a
c
c
Radiograph inspection of impeller Liquid penetrant inspection of impeller Liquid penetrant inspection of casing repairs Impact test Hydrostatic test Performance test Certified performance test and NPSH curve Mechanical run test with vibration measurements Manufacturer's standard mechanical run test Noise level test Dismantling in case of failed mechanical run test Balancing of rotating assembly Hardness test for impeller and casing welds*
Key: I = Centrifugal utility pumps < 50 hp or < 250 psig discharge pressure 2 = All centrifugal pumps > 50 hp or > 250 psig discharge pressure 3 = Process pumps < 50 hp or < 250 psig discharge pressure a = Visual inspection by company's inspector b = Certified records examined by company representative c = Certificates and data supplied by vendor *If HzS is present in fluid
142
Inspection and Testing
143
The buyer must determine the inspection procedures he or she wants. If a witnessed test is necessary, the inspector should be instructed to perform the following tasks: • Visually inspect castings • Check welding procedures • Visually inspect all parts before they are assembled. The castings shall not show signs of repair by peening, plugging, or brazing. • Check running clearances on vendor' s inspection records • Check impeller and rotor balance as recorded on vendor' s inspection records • Check pump package dimensions • Witness tests
Hydrostatic Test Hydrotest the pump casing at 1.5 times its design pressure. The gaskets used for this test should perfectly match the gaskets used in the pump. Leaks past the gaskets are not acceptable. Discard the test gaskets after the test. If any leakage or pressure drop occurs during the test, repeat the test after making necessary repairs. Test pumps whose medium has a specific gravity of 0.7 or less with kerosene. Use dye penetrant solution for the hydrotest if the casing pattern is new or changed, and/or if the pump media will consistently remain above 500*F. Any dyepenetration test shall adhere to ASTM E165. If the buyer does not require a witnessed hydrotest, the vendor shall submit its test data for client review.
Performance Test The buyer may insist on a performance string test, which shall include the actual driver, instrumentation, and eventual gear (see Figure 19.1 for setup). Because the vendor's lack of the electric power required may render this test inconvenient or impossible, most companies allow the vendor to use a shop driver. The shop driver often consists of a hydraulic turbine instead of an electric motor. Preferably, the test shall run at rated speed, but the vendor may run the test at other speeds. In this case, the vendor should submit the conversion-to-rated-speed calculations to the buyer for approval before starting the test. The actual string test will then occur at start-up. The manufacturer usually will use water during the performance test. If the real pump medium is viscous, the manufacturer shall make the necessary correction, according to the latest issue of the Hydraulic Institute Complete Pump Standards. The vendor often submits a list of the test equipment it will use. Upon request, the vendor shall submit calibration data for this equipment. Typical instruments include: • An orifice with a differential pressure gauge to measure the flow rate • Digital tachometer to measure pump rpm
144 •
• • • •
Practical Introduction to Pumping Technology
Double-element watt meter to measure power Pressure gauges to measure pump head Vibration monitor Temperature transmitter and RTD probe Sound-level meter
The performance test shall be combined with the mechanical run test, which shall last a minimum of four hours. The inspector shall check the following data, which should match the actual performance curve: • Head against flow • Power requirements at different flow rates • Efficiency at different flow rates The buyer shall select at least five flow rates for inspection. These should include the following points: • • • • • •
Shut-off Minimum flow as specified by vendor A point between rated and minimum flow Flow at best efficiency point Rated flow 110 percent of rated flow At each point, the inspector shall check the following:
• • • • • • •
Pump shaft vibration Shaft axial displacement Suction pressure Discharge pressure Flow rate Driver input power Noise level
If there is a gear in the pump package, the inspector shall check the gear contact pattern after the conclusion of the performance test. Gear contact shall be consistent. A contact area of less than 75 percent will affect gear performance. After the performance test, it is customary to open multistage and barrel pumps to inspect the rotor assembly. After draining the test liquid, the test crew shall dry all the pump components.
Inspection and Testing
145
ORIFICE
f
ISOL~TION
FLOW
VALVE
VAL
COl~Prl~O
r-
A
TURBINE
CLUTCH
MOTOR
GEAR
PUMP
Figure 19.1 Typical Performance Test Setup
NPSH Test It is a good idea to run a NPSH test on pumps 100 hp and larger where the NPSHA is close to the pump's required NPSH. The suppression test may determine the NPSH at 3 percent differential head loss or at 0 percent loss if the pump is in critical service with variable flow. The NPSH shall be plotted at the following flow rates, as a minimum: • • • •
Minimum continuous flow Midway between minimum flow and rated flow Rated flow 110 percent of rated flow